Yes, ethics approval for systematic reviews varies: published, non-identifiable work usually needs no REC/IRB; identifiable or contact-based work may.
Researchers ask this question when planning a literature synthesis or preparing a protocol. The short answer hinges on what kind of material you handle and whether your work counts as research with human participants under policy. Most library-only projects draw on published reports and never touch personal information. Other projects request datasets, pool individual-level records, or message authors and participants. That shift changes the oversight you need and the wording journals expect in your methods.
Quick Rule Of Thumb
Work that only summarizes published, non-identifiable data usually proceeds without a formal application to a Research Ethics Committee (REC) or Institutional Review Board (IRB). The moment you access identifiable records or interact with living people, you may trigger review. National regulators and journal policies align on this split, even if local procedures differ.
When Review Is Or Isn’t Needed: Common Scenarios
The matrix below shows typical cases you’ll meet during planning and write-up.
| Scenario | Data Or Activity | Typical Oversight |
|---|---|---|
| Published-only synthesis | Aggregate results from articles, trial registries, reports | No REC/IRB in most settings; keep a clear ethics statement |
| Contacting study authors | Email for clarifications or missing group-level numbers | Usually no REC/IRB; keep correspondence and consent to share clarifications |
| IPD meta-analysis | Individual participant data with identifiers removed or coded | Data-sharing agreements; REC/IRB may apply at source or your site |
| Unpublished clinical records | Access to charts, EHR exports, registries with identifiers | REC/IRB review expected; data controller approvals required |
| Social media scraping | User posts or profiles where identity can be inferred | High risk; many institutions require REC/IRB decision |
| Surveys or interviews | New contact with participants for missing outcomes | REC/IRB review required in most jurisdictions |
What The Major Guides Say
Cochrane’s methods text treats literature-only syntheses as secondary research and centers transparency, bias control, and data management. Journal guidance from medical editors points in the same direction: if people are involved or identifiable information is handled, include an ethics statement with committee name and reference number; if no approval was required, say so and state why. UK regulators add a clear tool for projects run in health settings to check whether NHS Research Ethics Committee review applies, and U.S. policy sets the threshold around interaction with living individuals or use of identifiable private information.
Two authoritative anchors worth bookmarking during planning are the UK REC decision tool and the U.S. Common Rule in 45 CFR 46. These pages define when a project involves a living person or identifiable private information, which is the line many committees and journals use.
Ethics Approval For Literature Syntheses: What Editors Expect
Editors want two things: the right oversight for the design you choose, and a clean statement that documents it. BMC’s Systematic Reviews journal, for instance, asks authors who used human data or tissue to name the committee and provide a reference number, and if approval was waived, to say so explicitly. Leading editors under the ICMJE umbrella also ask for clear reporting and structured abstracts for syntheses.
Cases Where No REC/IRB Is Usually Needed
Most evidence syntheses use only published sources: journals, trial registries, dissertations, and agency reports. In these designs you do not record new information from people and you do not process identifiable private information. That places the work outside human-subjects review in many regions. UK health regulators point researchers to their tool to confirm this for NHS settings, and U.S. policy places the line at interaction with living individuals or handling identifiable private information.
Typical cases include a review that extracts risk ratios from published trials, a meta-analysis of odds ratios from observational studies, or a methods review that compares risk-of-bias tools. None of these designs contact patients or read medical charts. You still need good practice: protocol registration, documented screening, dual data extraction, and a plan for bias assessment.
Cases That Can Trigger Ethics Review
Some projects under the “review” label cross into direct or indirect handling of human information. Here are common triggers:
Individual Participant Data (IPD) Pools
When you request row-level datasets, you often sign data use agreements and show your site’s approval or exemption letter. Even when identifiers are removed, re-identification risk may exist, which is why committees or data custodians ask for safeguards.
Use Of Unpublished Records
Extracts from hospital systems, insurance claims with identifiers, or local registries with personal fields usually need REC/IRB review. You also need approvals from the data controller or privacy officer.
Direct Contact With People
Follow-up surveys or interviews to fill gaps create interaction with living individuals, which can meet the threshold for human-subjects research under the Common Rule and similar policies. Plan consent, privacy notices, and secure storage.
Public Online Content
Posts on social platforms may look public, yet many institutions treat them as human data when individuals are identifiable. Expect an ethics decision before collection, especially if scraping at scale.
Practical Tests You Can Apply
Test 1: Is Anyone Interacting With You?
If you send a questionnaire, schedule an interview, or collect new outcome data from people, you are interacting. That step usually triggers review.
Test 2: Can A Person Be Identified?
If the dataset includes names, contact details, dates linked to small groups, face photos, or any unique codes that can map back to a person, you are in human-data territory.
Test 3: Did A Data Controller Approve Access?
When a hospital, registry, or company provides files, the custodian will expect an approval letter, a waiver, or a formal agreement. Without those, journals will flag the submission.
How To Document Your Ethics Position
Every manuscript benefits from a short, direct paragraph in the methods. Use plain words, avoid jargon, and be specific about what you did and did not do. Editors dislike vague claims. They want the committee name, a reference number when available, and a reason when no approval was required.
Template Language You Can Adapt
Library-Only Review
“This review used published, aggregate data and involved no interaction with individuals or use of identifiable private information; ethics committee review was not required under institutional policy.”
IPD Meta-Analysis
“We analysed de-identified participant-level datasets obtained under data use agreements. The University REC approved the protocol (Ref. 2025-123) and confirmed that consent procedures at the original sites covered secondary analysis.”
Author Queries Only
“We contacted corresponding authors for clarification of group-level results. No new human data were collected. The project fell outside human-subjects review per institutional guidance.”
Paperwork You May Be Asked For
Even when a committee application is not needed, editors and data custodians often ask for proof of care. Keep the following items handy:
- A dated protocol with version number and any amendments.
- A registration record or public timestamp for your plan.
- Data management notes that explain how files are stored, who has access, and how long archives are kept.
- For IPD projects, de-identification steps, a data dictionary, and signed data use agreements.
- For clinical sources, confirmation that the providing site reviewed the request and cleared sharing.
- If you ran a local ethics check, include the exemption or waiver letter.
Second Table: Journal And Policy Snapshots
| Source | What It Says | Implication For Authors |
|---|---|---|
| ICMJE recommendations | Structured abstracts and clear ethics reporting when human data are involved | State committee name/number or waiver; keep abstracts structured |
| BMC Systematic Reviews | Ethics approval and consent statement required for studies using human data | Add a methods paragraph naming the REC or waiver and consent status |
| UK HRA decision tool | Checker to see whether NHS REC review is needed for health projects | Run the tool; save the printable outcome in your files |
Good Practice Even When No REC/IRB Applies
- Register the protocol and keep a change log.
- Describe search strings, dates, and databases.
- Use two reviewers for screening and extraction when possible.
- Pre-specify outcomes and subgroup plans.
- Share reproducible materials: screening forms, code, and derivations.
- Cite data sources and respect licenses.
Common Pitfalls That Delay Publication
- No ethics paragraph at all, leaving editors to ask basic questions.
- Vague claims like “approved by the IRB” without a committee name or reference number.
- Collecting surveys mid-project without a prior decision from a committee.
- Uploading raw files with hidden identifiers in metadata.
- Skipping data use agreements for IPD, which leads custodians to block sharing.
Checklist You Can Run Before Submission
- Confirm whether you interact with people or handle identifiable private information.
- If yes, route the protocol through your local REC/IRB.
- If no, prepare a short statement that explains why approval is not required.
- When using IPD, secure data use agreements and a documented approval or exemption.
- Keep copies of committee letters, waivers, and tool outputs with your records.
- Match your statement to the target journal’s instructions for authors.
Why This Matters For Editors And Readers
A clear ethics path protects participants in source studies, respects data owners, and saves time during peer review. Readers gain confidence that the synthesis was planned, registered, and conducted with care. Committees and data controllers see that you recognise boundaries between published summaries and human information. That is the line policies draw, and it is the line your submission should respect.
Bottom Line For Authors
If your project only handles published, non-identifiable information, you usually don’t need a committee application. If people, charts, or identifiable records enter the picture, plan for oversight, approvals, and clear wording in the methods. State the decision plainly, show the paperwork when asked, and move on to the parts that readers care about: the question, the search, and the evidence. Clear wording speeds reviews and decisions. Now.